gpl.pierrick.brihaye.aramorph.lucene
Class ArabicStemAnalyzer

java.lang.Object
  org.apache.lucene.analysis.Analyzer
      gpl.pierrick.brihaye.aramorph.lucene.ArabicStemAnalyzer

public final class ArabicStemAnalyzer

extends org.apache.lucene.analysis.Analyzer

Analyzer for the arabic language. This analyzer uses Tim Buckwalter's algorithm (available at LDC Catalog) to identify the morphological category of arabic tokens. The significant grammatical categories are still to be determined but the current list gives good results. Final tokens are a romanized version of the canonical word.

Author:

Pierrick Brihaye, 2003

Field Summary

protected boolean

outputBuckwalter Whether or not the analyzer should output tokens in the Buckwalter transliteration system

Constructor Summary

ArabicStemAnalyzer()
Constructs an analyzer that will return grammatically significant arabic tokens in the Buckwalter transliteration system.

ArabicStemAnalyzer(boolean outputBuckwalter)
Constructs an analyzer that will return grammatically significant arabic tokens.

Method Summary

org.apache.lucene.analysis.TokenStream

tokenStream(java.lang.String FieldName, java.io.Reader reader) Returns a token stream of arabic words whose grammatically categories are found to be significant.

Methods inherited from class org.apache.lucene.analysis.Analyzer

tokenStream

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

outputBuckwalter

protected boolean outputBuckwalter

Whether or not the analyzer should output tokens in the Buckwalter transliteration system

Constructor Detail

ArabicStemAnalyzer

public ArabicStemAnalyzer()

Constructs an analyzer that will return grammatically significant arabic tokens in the Buckwalter transliteration system.

ArabicStemAnalyzer

public ArabicStemAnalyzer(boolean outputBuckwalter)

Constructs an analyzer that will return grammatically significant arabic tokens.

Parameters:

outputBuckwalter - Whether or not the tokens should be translitered

Method Detail

tokenStream

public org.apache.lucene.analysis.TokenStream tokenStream(java.lang.String FieldName,
                                                          java.io.Reader reader)

Returns a token stream of arabic words whose grammatically categories are found to be significant.

Parameters:

reader - The reader

Returns:

The token stream